Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/16382818

BoltonDC,McKinleyMP,PrusinerSB..Identificationofaproteinthatpurifieswiththescrapieprion.Science218:1309-1311

ARTICLEinSCIENCE·JANUARY1983

ImpactFactor:33.61·DOI:10.1126/science.6815801·Source:PubMed

CITATIONS

845

READS

60

3AUTHORS,INCLUDING:

DavidCBolton

NewYorkStateInstituteforBasicResearchi…

52PUBLICATIONS4,322CITATIONS

SEEPROFILE

Availablefrom:DavidCBolton

Retrievedon:24September2015

Identification of a Protein That Purifies with the Scrapie Prion

Abstract. Purification of prions from scrapie-infected hamster brain yielded aprotein that was not found in a similar fraction from uninfected brain. The proteinmigrated with an apparent molecular size of 27,000 to 30,000 daltons in sodiumdodecyl sulfate polyacrylamide gels. The resistance of this protein to digestion byproteinase K distinguished it from proteins of similar molecular weight found innormal hamster brain. Initial results suggest that the amount of this protein cor-

relates with the titer of the agent.

Scrapie is a transmissible, degenera-tive nervous system disease of sheep andgoats. Several features of scrapie closelyresemble two human diseases, kuru andCreutzfeldt-Jakob disease. Researchover the last two decades has demon-strated that the scrapie agent is extreme-ly resistant to physical and chemicaltreatments that inactivate bacteria, vi-ruses, and viroids (1, 2). Although theagent is known to be a small, hydropho-bic particle (2-4), its precise structureand biochemical composition have elud-ed definition. A scrapie-specific structur-al macromolecule has not yet been iden-tified.

Studies with substantially purifiedpreparations of the scrapie agent haveshown that a protein is required for in-fectivity. Based on these studies, a newterm, "prion," was introduced to de-scribe and identify these unusual in-fectious particles (2). The agent in thepurified preparations was inactivated byprotease digestion and by reversiblechemical modification with diethyl pyro-carbonate (DEP) (4, 5). In addition,the agent was inactivated by chemicalreagents that denature proteins; thesereagents included sodium dodecyl sul-fate (SDS), urea, guanidinium thiocya-nate, and phenol (2, 6-8). The mostlikely explanation for these observationsis that inactivation results from,modifica-tion of a functional protein within thescrapie agent.The agent was obtained from weanling

random-bred (LVG/LAK) hamsters thatwere given an intracerebral inoculationwith 107 ID50 (mean infectious dose)units of the scrapie agent (6). Sixty dayslater, the infected brains were collectedand homogenized. The agent was puri-fied by low-speed centrifugation, poly-ethylene glycol precipitation, enzyme di-gestion, ammonium sulfate precipitation,and sedimentation into a discontinuoussucrose gradient (25 to 60 percent) (9).Samples collected throughout the purifi-cation were analyzed for infectivity bythe incubation time interval assay (6, 10)and for protein concentration by the Pe-terson-Lowry method (11).

Infectivity analysis of discontinuoussucrose gradients (Fig. IA) showed thatapproximately 50'percent of the infec-SCIENCE, VOL. 218, 24 DECEMBER 1982

tious agent was found near the bottom ofthe gradient in fraction 2. This fractionrepresented the interface of the 25 and 60percent sucrose solutions. It was en-riched for the scrapie agent since most ofthe applied protein remained in the topportion of the gradient. This procedureproduced a 100- to 1000-fold purificationof the agent; the specific infectivities offraction 2 ranged from 3.8 x 109 to1.2 x 10" ID50 units per milligram ofprotein (9).

Analysis of 125I-labeled sucrose gradi-ent fractions by gel electrophoresis re-vealed a diffuse protein band with amolecular size of 27,000 to 30,000 dal-tons in those fractions enriched for theinfectious agent (Fig. 1B). This band wasreadily detected in fraction 2 and to alesser extent in fractions 1 and 3. It wasalso present in fraction 15, but was diffi-cult to visualize because of the high-concentration of contaminating proteins'.The presence of this protein in frac-

tions containing the substantially puri-fied agent suggested there might be a

9.0

8.5E

o 8.0C]

S. 7.5a)

o' 7.0-6

6.5

2 6 10 14B

2 6 10Fraction number

14

0.4 E

E

0.3 o.__

co

a

C

a

specific relation between this protein andthe infectious particle.'For this reason,we attempted to determine whether or'not the protein was present in fractionsanalogously purified from brains o'f age-matched,' uninoculated hamsters. Sam-'ples of fraction 2 from six different prep-arations of scrapie-infected hamsterbrain and three different preparations ofnormal hamster brain were' I-labeledand analyzed by SDS polyacrylamide'gelelectrophoresis.'The protein of interestwas found in all the &amples from scra-`pie-infected brains but appeared to beabsent from samples of normal' prains(Fig. 2). These observations were' com-plicated by the pfesence of discrete pro-tein bands that migrated w?ithin the27,000- to 30,000-dalton region in somescrapie and normal samples (Fig. 2, lanes'4 to 8).

Subsequent investigations showedthat the protein was readily degraded'by'proteases if it was denatured prior todigestion. After denaturation at '100°C inthe presence of 1.25 percent SDS and1.25 percent ,B-mercaptoethanol, the pro-tein was hydrolyzed by proteinase K,pronase, and Staphylococcus aureus V-8protease. These results confirmed thatthis macromolecule is indeed a protein.Attempts to digest the protein prior to

denaturation were not successful. In-stead, we found that this protein' wasresistant to digestion by proteinase:.Kunder nondenaturing conditions.-Ftac-tions prepared 'from scrapie' and rormalhamster brains were '251-labeled' withBolton-Hunter reagent (Fig. 3A) or'' C-labeled DEP (Fig. 3B) and incubatedwith proteinase K. The electrophoreticprofiles of those samples for'which the

3.2 .@ Fig. 1. Scrapie infectivity and '251-labeled* proteins in sucrose gradient fractions. The.

partially puiified scrapie agent (4 ml in 20 miM. tris-acetate, 1 mM EDTA, 1 mM dithiothrei-:o tol, 0.2 percent Sarkosyl, 2 percent TX-100':

and 0.8 percent SDS, pH 8.3) was layered ona discontinuous sucrose gradient consisting of.60 percent sucrose (5 ml) and 25 percentsucrose (30.5 ml) in 20 mM tris-acetate, i 'aikEDTA, pH 8.3, containing no detergents. 1Theagent was sedimented into the gradient at50,000 rev/min (Beckman VTi 50 rotor) for 2hours at 4°C. Gradient tubes were fractionat-ed from the bottom (fraction 1) into 15 sam-ples of 2.5 ml each (9). (A) Infectious scrapietiters (0) and protein concentrations (A) weredetermined for each fraction (6, 10, 11). (B)Portions of each gradient fraction were con-

* centrated tenfold and labeled with N-succin-imidyl 3-(4-hydroxy, 5-['251]-iodophenyl)pro-pionate as described (9, 13, 14). The sampleswere subjected to electrophoresis through a 5to 20 percent linear gradient polyacrylamidegel (15). Radioautographic exposure of thedried gel was performed at room temperaturefor 1.5 hours.

0036-8075/82/1224-1309$01.00/0 Copyright © 1982 AAAS 1-309

on

May

7, 2

007

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

proteinase K was omitted are shown inlanes 1 to 4. Digestion with proteinase Kprior to electrophoresis removed most ofthe labeled bands from both scrapie andcontrol fractions; however, the resistantprotein described above remained intactin scrapie samples (lanes 5 and 6). Theprotein was not found in control samplesprepared from uninfected hamster brains(lanes 7 and 8). Although the molecularbasis for the resistance of this proteinto degradation by proteinase K is un-known, we presume it results from theprimary structure of the molecule. Ininitial studies, one-dimensional peptidemaps have shown that the protein ofinterest and the proteins of similar mo-lecular size from uninfected hamsterbrain generate distinctly different pep-tide fragments (12).

In addition to the date presented inFigs. 2 and 3, we have confirmed thepresence of the protease-resistant pro-tein in purified fractions from six otherpreparations of scrapie-infected hamsterbrain. The protein was not found inanalogous fractions from five other prep-arations of normal hamster brain, includ-ing four preparations made from thebrains of animals which were inoculated

with normal hamster brain and killedapproximately 60 days later. Since thisproteinase K-resistant protein has notbeen found in purified fractions fromnormal hamster brains, we conclude thatthe protein is specifically associated withscrapie infection.Our data provide evidence for a pro-

tein which is associated with scrapieinfection. There are at least two hypoth-eses that satisfactorily explain our obser-vations: (i) the scrapie-associated pro-tein may be a pathological product ofscrapie infection, or (ii) it may be astructural component- of the scrapieagent. If the first hypothesis were true,several different models could explainthe appearance of a specific protein inthe brain during the course of the dis-ease. First, a pathological process, suchas astrocyte proliferation, might increasethe synthesis of a protein normally pres-ent in the brain at a concentration belowour limit of detection. Second, scrapieinfection could induce the synthesis of aprotein which is not normally expressedpostnatally in the brain. Third, genera-tion of this protein from an existing brainprotein might occur as a consequence ofproteolytic degradation, glycosylation,

A i 2 3 4 5 6 7 8

93

es

45

31

2214

Fig. 2 (left). Analysis of scrapie and normalfraction 2 by gel electrophoresis. Portions of s S N N S S N N

fraction 2 from six different scrapie-infectedbrain preparations (lanes I to 6) and three

different normal brain preparations (lanes 7 to

9) were subjected to electrophoresis in a 5 to20 percent linear gradient polyacrylamide gel

after I251-labeling as described in Fig. 1. Ra-dioautographic exposure of the dried gel was

performed at room temperature for 30 min-

utes; lanes 1, 2, and 9 were enhanced byexposure for 20 hours. Fig. 3 (right). Plro-teinase K digestion of concentrated scrapie

and normal fraction 2 preparations. Concen-

trated and labeled samples of scrapie (S) andnormal (N) fraction 2 were incubated for 30

minutes at room temperature in 10 mM tris-Cl, pH 7.4,(lanes I to 4) or the same buffercontaining proteinase K (100 ,ug/ml) (lanes 5

to 8). The digestion was terminated by the addition of an equal volume of 2 x concentrated SDSelectrophoresis sample buffer and heating to 100°C for 2 minutes. Electrophoresis wasperformed in 15 percent polyacrylamide gels by the Laemmli method (15). (A) Samples were

'25I-labeled as described above. (B) Samples were labeled in 60 mM sodium phosphate, 0.2percent Sarkosyl (pH 7.2) with 10 mM 14C-labeled DEP (53 mCi/mmole). The reaction was

performed at room temperature for I hour. The gels were processed for fluorography by themethod of Bonner and Laskey (16).

1310

fatty acid acylation, or other chemicalmodification induced by scrapie infec-tion. Although these models could ac-count for the appearance of such a pro-tein in scrapie-infected brains, they pro-vide no obvious explanation for the co-purification of this protein and thescrapie agent.The second hypothesis, that this pro-

tein is a structural component of thescrapie agent, is supported by severallines of evidence. The protein has beendetected in all highly purified scrapieagent fractions examined and was shownto be distinct from purified normal brainproteins. The physical and chemicalproperties of the associated protein aresimilar to those of the scrapie agent, asdemonstrated by their copurificationthrough a series of procedures whichresult in a specific enrichment for theagent (9). In our purest fractions, thisprotein accounted for as much as 10percent of the total protein. The proteincan be isotopically labeled by DEP,which has been shown to react with acomponent of the agent, thereby causinginactivation (7).

Resistance of this protein to protein-ase K-catalyzed degradation may appearto be inconsistent with the sensitivity ofthe agent to this enzyme (4). However,significant inactivation of the agent byproteinase K was observed only afterincubation at 37°C for three or morehours (4). Thus, the resistance of thisprotein to proteinase K digestion is con-sistent with the behavior we would pre-dict for a component of the scrapieagent. Further investigations will berequired to conclusively demonstratewhether or not the protein describedabove is required for the infectivity ofthe agent.

DAVID C. BOLTONMICHAEL P. MCKINLEYSTANLEY B. PRUSINER

Departments of Neurology andBiochemistry and Biophysics, Universityof California, San Francisco 94143

References and Notes

1. G. D. Hunter and G. C. Millson, J. Gen. Micro-biol. 37, 251 (1964); J. Comp. Pathol.77, 301 (1967); I. H. Pattison, ibid. 75, 159(1965); T. Alper, D. A. Haig, M. C. Clarke,Biochem. Biophys. Res. Commun. 22, 278(1966); T. Alper, W. A. Cramp, D. A. Haig, M.C. Clarke, Nature (London) 214, 764 (1967); G.D. Hunter, S. C. Collis, G. C. Millson, R. H.Kimberlin, J. Gen. Virol. 32, 157 (1976); D. C.Gajdusek, Science 197, 943 (1977); T. 0. Diener,M. P. McKinley, S. B. Prusiner, Proc. Natl.Acad. Sci. U.S.A. 79, 5220 (1982).

2. S. B. Prusiner, Science 216, 136 (1982).3. S. B. Prusiner et al., Biochemistry 17, 4993

(1978).4. S. B. Prusiner et al., Proc. Natl. Acad. Sci.

U.S.A. 78, 6675 (1981).5. M. P. McKinley, F. R. Masiarz, S. B. Prusiner,

Science 214, 1259 (1981).6. S. B. Prusiner et al., Biochemistry 21, 4483

(1980).

SCIENCE, VOL. 218

on

May

7, 2

007

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

7. G. C. Millson and E. J. Manning, in SlowTransmissible Diseases of the Nervous System,S. B. Prusiner and W. J. Hadlow, Eds. (Aca-demic Press, New York, 1979), vol. 2, p 409.

8. W. B. Melchior and D. Fahrney, Biochemistry9, 251 (1970); J. J. Holbrook and V. A. Ingram,Biochem. J. 131, 729 (1973); R. B. Wallis and J.J. Holbrook, ibid. 133, 183 (1973); Y. Burnstein,K. A. Walsh, H. Neurath, Biochemistry 13, 205(1974).

9. S. B. Prusiner et al., Biochemistry, in press.10. S. B. Prusiner et al., Ann. Neurol. 11, 353

(1982).11. G. L. Peterson, Anal. Biochem. 83, 346 (1977).12. D. C. Bolton, M. P. McKinley, S. B. Prusiner,

in preparation.13. R. K. Durbin, thesis, University of California,

Davis (1982).

Increases in corticosteroid concentra-tions are usually considered to be medi-ated by the pituitary-adrenal axis and toresult from increased release of adreno-corticotropin (ACTH) from the pituitarygland (1). Recently, we demonstrated invitro that substances that induce inter-feron (IFN)-a, such as viruses, causeproduction of an ACTH-like substancefrom human lymphocytes (2) that isstrongly associated with IFN-a. TheACTH-like substance and IFN-a weredissociated by acid (pH 2) and werenever associated if lymphocytes wereinduced in the presence of tunicamycin(2). These findings strongly suggest that

14. A. E. Bolton and W. M. Hunter, Biochem. J.133, 529 (1973).

15. U. K. Laemmli, Nature (London) 227, 680(1970).

16. W. M. Bonner and R. A. Laskey, Eur. J.Biochem. 46, 83 (1974).

17. We thank Drs. A. Liebman and R. Kaback ofthe Roche Molecular Biology Institute for pro-viding the '4C-labeled diethyl pyrocarbonate andD. Groth, S. P. Cochran, K. Bowman, D. Dow-ney, J. Mayled, and N. Mock for technicalassistance. Supported by postdoctoral fellow-ship from the Muscular Dystrophy Association(to D.C.B.), by National Institutes of Healthresearch grants NS14069 and AG02132, and byR. J. Reynolds Industries, Inc.

6 July 1982; revised 19 October 1982

the ACTH-like substance and IFN-aare coordinately induced products of dif-ferent genes and become associatedthrough a carbohydrate interaction.These findings imply that certain stimuli,which effect lymphocyte production ofIFN, should not require the participationof the pituitary gland for an ACTH-mediated increase in corticosteroids.Furthermore, they suggest that lympho-cyte production of ACTH in vivo mightbe dissociable from IFN production.These concepts were tested by determin-ing (i) whether virus-infected hypophy-sectomized mice would have an increasein ACTH-like activity, thus showing an

elevated cqrticosterone concentrationand (ii) whether this putative response isdissociable from IFN production.Hypophysectomized, female Swiss

Webster mice were infected with New-castle disease virus (NDV) (3, 4). Thesemice showed a time-dependent increasein corticosterone production (Fig. lA).Peak ster'oid concentrations occurred at8 hours after infection and thereafterdeclined. Serum IFN increased coor-dinately with corticosterone (data notshown). Thus the production of cortico-sterone was related to the induction ofIFN and, by inference, induction of theACTH-like substance. Corticosteroneconcentrations at 8 hours (10 ± 2.6,mean + standard error, N = 13) differedfrom those at 0 hour (2.76 ± 0.11,N = 19) at a probability of < .01 (two-tailed Student's t-test). Whereas hy-pophysectomized animals did not re-spond to a classical "stressor" (coldwater) they did respond to ACTH (Fig.IB). The response to a pharmacologicdose of ACTH was not maximal sincethere was probably some adrenal atro-phy in the 5 days between the hypophy-sectomy and the experiment (4). Thissuggests that under optimal conditionsthe adrenal response to virus might beeven higher than the 3.5-fold increasethat was observed.The production of ACTH by the pitu-

itary gland, and thus the production ofsteroid by the adrenal glands, is con-trolled by negative feedback via adrenalcorticosteroids (1). The virus-induced in-creases in corticosterone concentrationin hypophysectomized mice appear tobe similarly controlled, because priortreatment with the synthetic steroiddexamethasone completely inhibited theincrease in corticosterone at 8 hoursafter NDV infection (Fig. 1B). That lym-phocytes are a likely source of an

A

T

0 2 4 6 8

Hours after NDVo

C.)

C-

Fig. 1. (A) The kinetics of corticosterone production after NDV infec-tion of hypophysectomized mice. Mice were injected intraperitoneallywith 800 hemagglutination (HA) units of NDV in 0.2 ml. At the indi-cated times, the mice were decapitated and trunk blood was collectedfor corticosterone determination by radioimmunoassay (7). Mouseheads were dissected and examined under a dissecting microscope, andanimals with any remnant of the pituitary gland were discarded. Thenumbers of mice for 0, 2, 4, 6, 8, 10, and 12 hours were 19, 2, 9, 16, 13,12, and 5, respectively. (B) Dexamethasone (dex.) suppression ofcorticosterone production by NDV-infected hypophysectomized mice.The mice were given free access to water with or without dexametha-sone (20 ,ug/ml) for approximately 24 hours. Control (NDV, N = 13) ordexamethasone-treated (NDV + dex, N = 5) mice were then injected,as before, with NDV. Eight hours after infection the mice were killedand treated as in (A). Other controls without dexamethasone included:controls (N = 19), in which saline (0.2 ml) was injected intraperitoneal-ly and the mice were killed 8 hours later; cold stress (N = 5), in whichthe mice were placed in ice water for 45 seconds and killed 20 minuteslater, and ACTH (N = 9), in which the mice received an intramuscularinjection of 25 ,g of Cortrosyn (Organon, Inc., West Orange, NewJersey) in 0.2 ml and were killed 2 hours later. Corticosterone concen-trations are expressed as means + standard error.

0036-8075/82/1224-1311$01.00/0 Copyright 1982 AAAS

Virus-Induced Corticosterone in Hypophysectomized Mice:A Possible Lymphoid Adrenal Axis

Abstract. Infection of hypophysectomized mice with Newcastle disease viruscaused a time-dependent increase in corticosterone and interferon production. Priortreatment with dexamethasone completely inhibited the virus-induced elevation incorticosterone concentration, but did not significantly alter the interferon response.

Lymphocytes appear to be the most likely source of an adrenocorticotropin-likesubstance that is responsible for the increased corticosterone, since spleen cellsfromthe virus-infected, but not from control or dexamethasone-treated, hypophysecto-mized mice showed positive immunofluorescence with antibody to adrenocorticotro-pin-(1-13 amide). Thus the adrenocorticotropin-like material and interferon appear

to be coordinately induced and differentially controlled products of different genes.

These findings strongly suggest the existence of a lymphoid-adrenal axis.

20 1

9

0i%B 8 -

0, 7 -c0

4-

000

00

4-

30

T;

SCIENCE, VOL. 218, 24 DECEMBER 1982 1311

on

May

7, 2

007

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om